Quick phonology question...

[Bristel]

(the first are underlying phonemes rounded due to a cluster labialization rule, the second are the surfaced allophones)
b̥ʷ > b̥͡β̥ʷ
pʷʰ > p͡ɸʷ
hʷ > ɸʷ
lʷ > ɫʷ
ɾʷ > ʁʷ
jʷ > ɥ

What kind of phonological thing is behind this? A consonant harmony rule then a lenition (or is it a kind of fortition?) for some of the consonants? (note that the consonants on the list are only a part of the phonemes that can be rounded, the rest do not have this issue).

Can /mʷ/ exist? What would it turn into if not?

(If Nortaneous could help out with this issue, I'd appreciate it, as he was the one to come up with the phonology I am using)

[Rui]

jʷ > ɥ

Not sure about the others, but these two are equivalent (i.e. "labialized palatal approximant")

[Bristel]

jʷ > ɥ

Not sure about the others, but these two are equivalent (i.e. "labialized palatal approximant")

Ah I see, I forgot that aspect.

[Rory]

b̥ʷ > b̥͡β̥ʷ
pʷʰ > p͡ɸʷ

This is lenition.

hʷ > ɸʷ

This is arguably a kind of fortition. It's similar to Northeastern Scots, where historical /hw/ shifted to /f/.

lʷ > ɫʷ
ɾʷ > ʁʷ

These are both a "backing" of some sort, and to be honest I can't see any phonetic motivation behind it. (Not that that is necessary.)

The bigger question, though, is what do you gain from having a name to apply to all of these? As conlangers we love classifying things, but it's only worth the effort if it's actually useful to us in. (Of course, sometimes the utility is simply in learning about linguistics, which is pretty rad.)

Can /mʷ/ exist? What would it turn into if not?

Think about it. Can you round your lips when they're closed? If so, does it change the sound produced? Nope. If you're wanting some "labialization" effect, maybe it could spread to adjacent vowels? But you can't make the labial more labial, I'm afraid.

[Bristel]

Why can I see phoneme inventories all over with /mʷ/? Is it really /mw/?

I gain from naming these things because I don't have a name for them. I am trying to describe my conlang for a wiki, and I don't have proper terms to describe the mutations occurring. Should I just call it mutation and leave it at that?

I could further describe the fact that the first two are actually lenition, and the rest are some mixed kinds of mutations.

[Nortaneous]

I don't think there's any general term for it. I doubt you'd even need to have a name for it; just say "these consonants are realized as this when rounded".

lʷ > ɫʷ
ɾʷ > ʁʷ

These are both a "backing" of some sort, and to be honest I can't see any phonetic motivation behind it. (Not that that is necessary.)

Rounding is generally associated with backing of some sort. IIRC there was some language where rounded alveolars shifted to retroflexes.

Can /mʷ/ exist? What would it turn into if not?

Think about it. Can you round your lips when they're closed? If so, does it change the sound produced? Nope. If you're wanting some "labialization" effect, maybe it could spread to adjacent vowels? But you can't make the labial more labial, I'm afraid.

Except <ʷ> is usually used for labiovelarization, so even ignoring the on/offglides that could be present, yes it can.

[Bristel]

Sure, I've thought about calling all of it "lenition", as the term is used differently within some specific language studies anyway.

Thanks for the answers Rory and Nort.

Actually, I knew they were equivalent, Chibi, but for purposes of explaining, that underlying phone comes from a situation in which the first syllable ends in a labialized consonant, and the next starts with /j/, the roundedness rule applies, which makes /j/ > /jʷ/ > /ɥ/.

what would have made it really confusing would be to represent /j w/ as <j w> but have /jʷ/ represented as <jw> as would have been indicated by the original orthography.

Now, /jʷ/ (/ɥ/) is represented as <jo>, with a letter that is unused for vowels sounds. (there's no /o/ in Teskwan)

Let's make a clearer example, two separate words then a compound of the two:

puko /pʰukʷʰ/
jat /jatʰ/
pukjoat /pʰukʷʰjʷat/ [pʰukʷʰɥat]

(<o> is inserted after the whole cluster, not after each rounded consonant)

[finlay]

Can /mʷ/ exist? What would it turn into if not?

Think about it. Can you round your lips when they're closed? If so, does it change the sound produced? Nope. If you're wanting some "labialization" effect, maybe it could spread to adjacent vowels? But you can't make the labial more labial, I'm afraid.

Except <ʷ> is usually used for labiovelarization, so even ignoring the on/offglides that could be present, yes it can.

hey nort, which of you and rory is studying phonetics at phd level? thought so. the ipa says "labialization", and it really doesn't matter at all if you've seen one or two people using the symbol for something different, in that respect. (it also really drives me up the wall when you say that [ɹ] doesn't stand for an alveolar/coronal approximant, which it does. you are welcome to use it for "bunched R" without explaining what "bunched R" is, and that is your prerogative, but that's not the standard usage)

that said, on a phonological level it is likely that labialization and velarization go somewhat hand in hand, and i'm fairly sure i've seen languages with /mʷ/ out there.

[Hakaku]

hey nort, which of you and rory is studying phonetics at phd level? thought so.

What does this have to do with anything? The Handbook of the International Phonetic Association (1999), by the International Phonetic Association, explicitly cites Amharic as having the labialized consonants /fʷ/, /bʷ/, /mʷ/, /pʷʼ/, /tʷʼ/, /kʷ/, /kʷʼ/, /hʷ/. It also goes on earlier to state that "labialized" is used very broadly, and that velarization is often a co-existing non-contrastive property, which is why there's no reason to overly mark it and that the symbol /ʷ/ is most appropriate. Nortaneous' wording was not incorrect, you're just trying to nitpick a fight.

[Nortaneous]

And there's an ExtIPA diacritic for "open rounding", which, as far as I can tell, is supposed to be rounding without velarization.

[Rory]

The open-rounding diacritic is used to mean, as far as I understand, rounded lips that are relatively open - as opposed to the more pursed lips of the superscript w. Bear in mind that ExtIPA is primarily intended to be used for the transcription of disordered speech. Also bear in mind that the very concept of phonetic transcription is based on modes of thought from the 1800s. (On this subject, I highly recommend reading Munson et al's "Deconstructing phonetic transcription: Language-specificity, covert contrast, perceptual bias, and an extraterrestrial view of vox humana." Clinical Linguistics and Phonetics, 24: 245-260.)

Also, I'm afraid that "studying phonetics at the PhD level" does not entail learning the minutiae of IPA usage. Still, I don't understand why one would want to transcribe a velarized labial nasal as [mʷ] rather than [mˠ].

On the subject of lip-rounding being correlated with backing, this is true. Lip-rounding effectively extends the length of the vocal tract, and thus lowers the resonant frequencies. A similar effect can be done with tongue-backing, to the extent that if a speaker is unable to round their lips for whatever reason, they will draw their tongue back further. (Perkell, J.S., Matthies, M.L., Svirsky, M.A. and Jordan, M.I. (1993). Trading relations between tongue-body raising and lip rounding in production of the vowel /u/: A pilot motor equivalence study. Journal of the Acoustical Society of America 93: 2948-2961.)

[Radius Solis]

The thing with [ʷ] is that it is not just a simultaneous co-articulation. If you want I can find you the passage in Sounds of the World's Languages to back this up later, but the gist of it is that labialized consonants like [kʷ] are not known to occur without some degree of a [w]-glide after the stop's release (or, rarely, before its closure). Thus the labels [kʷ] and [kw] have no sharp articulatory difference - and the reasons one might prefer one label over the other are frequently phonological instead of phonetic anyway.

So in keeping with that, transcribing [mw] as [mʷ] doesn't seem unreasonable when it appears as part of a systematic labialization distinction.

[jal]

labialized consonants like [kʷ] are not known to occur without some degree of a [w]-glide after the stop's release (or, rarely, before its closure). Thus the labels [kʷ] and [kw] have no sharp articulatory difference

I have always assumed that labialized consonants have rounding as well, which may not be present in a sequence of consonant + /w/. The w-glide is then a logical consequence of the rounding.


JAL

[Burned_Toast]

If you're looking for a language that uses labialised bilabial consonants then look up Arrernte (http://en.wikipedia.org/wiki/Arrernte_language) or look at the numerous Papuan New-Guinea languages because a fair few of those languages happen to use labialised bilablial consonants.

But you may be looking for an IE language. Take a look at Adyghe then. Something a bit closer to home? I remember that a dialect of Irish Gaelic uses /bʷ, pʷ, mʷ/ but they're only in minimal pairs.

As for /hʷ > ɸʷ/ It makes sense that /h/ would collapse to /ɸ/ because of lenition but then again, I've never seen the phoneme /ɸʷ/ in a natlang before. I actually think that this would just change to /ʍ/ or to /w/.

Also I've never heard of /p͡ɸʷ/ ever occurring labialised but I think that this phoneme wouldn't occur because the /ɸ/ would be erased and it would just be /pʷ/ but that's just a guess.

Can /mʷ/ exist? What would it turn into if not?

Think about it. Can you round your lips when they're closed? If so, does it change the sound produced? Nope. If you're wanting some "labialization" effect, maybe it could spread to adjacent vowels? But you can't make the labial more labial, I'm afraid.

You do realise that it's called secondary articulation for a reason, right? The primary consonant is pronounced and then the secondary consonant - which, in this case is /w/ - is pronounced as if it were a consonant cluster such as /st/. Just to put it simply. Refer to the top if you want example natlangs.

[finlay]

You do realise that it's called secondary articulation for a reason, right? The primary consonant is pronounced and then the secondary consonant - which, in this case is /w/ - is pronounced as if it were a consonant cluster such as /st/. Just to put it simply. Refer to the top if you want example natlangs.

I think you're missing the point a little bit – clusters and secondary articulations are different, and in an ideal world, secondary articulations are made simultaneously with the primary articulation, rather than one after the other. Labialisation is normally viewed as rounding the lips as you produce the primary articulation – as others have pointed out, it's not always that simple, and a labialised labial is probably used in phonology with another phonetic meaning, such as velarisation or consonant cluster with [w], but Rory is correct in saying that true labialisation (ie lip rounding as you produce the primary articulation) is impossible with bilabials. Radius is also correct in saying that even so, there's no sharp difference between the secondary articulation and the consonant cluster. That doesn't mean they aren't different, though. It's where theoretical phonetics starts to diverge from practical phonology; what we get taught in phonetics classrooms doesn't necessarily match up to how the things are used in reality.

To take another example, think of [c], a palatal plosive, created by raising the front of the body of your tongue towards the palate and touching it, and making a closure so that no air can escape until you release your tongue – now, bearing in mind that ʲ refers to palatalisation, ie simultaneously raising the front of the body of your tongue towards the palate, but keeping away from it to some degree, you may be able to see that [cʲ] is both tautological and impossible. [cj], of course, is very possible, where you have the [j] articulation keeping away from the palate right after the [c] articulation touching it. Phonologically, however, it may be beneficial for us to use /cʲ/, especially if there's a pattern in the language, and in practice, there isn't a sharp difference (as radius says) between [kʲ], with the [j] articulation simultaneous with the [k], and [kj], with it coming strictly after the [k]. There will be overlap.

[Bristel]

What is the phoneme in Swahili mwera 'hen, female bird', or the Bantu language Mwera, if not /mʷ/? Is it a sort of syllabic /m/ followed by /w/?

[Drydic]

What is the phoneme in Swahili mwera 'hen, female bird', or the Bantu language Mwera, if not /mʷ/? Is it a sort of syllabic /m/ followed by /w/?

Why can't it just be [mwera]? There's no syllabicity needed.

[jal]

Rory is correct in saying that true labialisation (ie lip rounding as you produce the primary articulation) is impossible with bilabials.

I'm not convinced. I can pronounce a rounded bilabial just fine, producing a sound slightly different from the non-rounded version. But I think the aforementioned offglide (or, in case of a trailing rounded vowel the lack of an "onglide") accounts for the largest acoustic difference.


JAL

[Davoush]

I may have missed something and risk sounding ridiculous but I just tried to pronounce [m] and [m_w], when I made [m_w] I noticed my mouth was in the position as if about to kiss (think of 'mwah') - does that count as rounding?

[AnTeallach]

I may have missed something and risk sounding ridiculous but I just tried to pronounce [m] and [m_w], when I made [m_w] I noticed my mouth was in the position as if about to kiss (think of 'mwah') - does that count as rounding?

That's what I thought too.

[jal]

does that count as rounding?

That is rounding. Saying [w], do your lips really make a different shape then with [m_w]?


JAL